R help - Nov 2008 - Estimating the standard error when you have sampling weights.

Hi,

Where can I find information ( freely available on the Internet , and also
books or other sources ) on how having sampling weights changes the
calculation of the standard error (of means and proportions)?

How good is R for this type of procedure? And SAS?

thanks

Robert

	[[alternative HTML version deleted]]

I am having difficulty thinking that you cannot find general material  
by doing a Google search, but can tell you from memory that the US  
National Center for Health Statistics publishes on the WWW quite a bit  
of information about their survey methods.

For an R-centric answer: Have you looked at the survey package that  
Lumley created?

Doing  help.search("sampling") I also see that wtd.mean is available  
in Harrell's Hmisc. The help page for that function also has useful  
links.

-- 
David Winsemius

On Nov 24, 2008, at 3:34 PM, Robert Wilkins wrote:
> Hi,
>
> Where can I find information ( freely available on the Internet ,  
> and also
> books or other sources ) on how having sampling weights changes the
> calculation of the standard error (of means and proportions)?
>
> How good is R for this type of procedure? And SAS?
>
> thanks
>
> Robert
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

On Mon, 24 Nov 2008, Robert Wilkins wrote:
> Hi,
>
> Where can I find information ( freely available on the Internet , and also
> books or other sources ) on how having sampling weights changes the
> calculation of the standard error (of means and proportions)?
>
Alan Zaslavsky keeps a comprehensive list of software for complex surveys, 
at
http://www.hcp.med.harvard.edu/statistics/survey-soft/

Although I'm biased, I think the 'survey' package in R is better
than
either SPSS or SAS for this purpose. The main competition would be Stata 
and the specialised packages such as SUDAAN and WesVar.

If you just want means and proportions, though, any of the software would 
be perfectly adequate.

The PEAS project at Napier University has some nice introductory material, 
although some of their software comparisons are a bit out of date:
http://www.napier.ac.uk/depts/fhls/peas/

 	-thomas

Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle